Nearly Optimal State Feedback Control of Constrained Nonlinear Systems Using a Neural Network HJB Approach

Abstract

In this paper, we treat constrained optimization of nonlinear systems. A rigorous solution method to obtain nearly optimal state feedback control that takes into consideration, actuator saturation, state space constraints, and minimum-time control requirement is presented. The constraints are encoded into the optimization formulation through special nonquadratic functionals. The associated Hamilton-Jacobi-Bellman (HJB) equation is then solved successively. Nonlinear approximating networks are used to obtain an approximate closed form solution of the value function of the HJB equation, which is then used to obtain a state feedback controller. The solution is carried over a compact set of the asymptotic stability region of an initial stabilizing control.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 2003
Accession Number
ADA417841

Entities

People

  • Frank L. Lewis
  • Murad Abu-khalaf

Organizations

  • University of Texas at Arlington

Tags

Communities of Interest

  • Weapons Technologies

DTIC Thesaurus Topics

  • Abstracts
  • Algorithms
  • Artificial Intelligence
  • Computer Programming
  • Control Systems
  • Control Theory
  • Differential Equations
  • Equations
  • Feedback
  • Linear Differential Equations
  • Linear Systems
  • Lyapunov Functions
  • Neural Networks
  • Nonlinear Systems
  • Optimization
  • Partial Differential Equations
  • Riccati Equation

Fields of Study

  • Mathematics

Readers

  • Mathematical Modeling and Probability Theory.
  • Operations Research
  • Robotics and Automation.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Space
  • Space - Spacecraft Maneuvers